Estimation of Genetic Parameters for Direct, Maternal and Grandmaternal Genetic Effects for Birth, Weaning and Six Month Weights of
Hanwoo (Korean Cattle)
S. B. Choi*'* 12, J. W. Lee1, N. S. Kim2, S. H. Na, J. F. Keown1 and L. D. Van Vleck3 National Livestock Research Institute, RDA, Cheonan 330-800, Korea
* Address reprint request to S. B. Choi. Tel: +82-417-580- 3452, E-mail: csb3452@animal.nlri.go.kr.
1 Department of Animal Science, University of Nebraska, Lincoln, NE 68583-0908, USA.
즈 Department of Animal Science, Chungbuk National University, Chongju 360-800, Korea.
3 USDA, ARS, Roman L. Hruska U.S. Meat Animal Reseaich Center, Lincoln, NE 68583-0908, USA.
Received February 6, 1999; Accepted May 28, 1999
ABSTRACT : The objectives of this study of Hanwoo (Korean Cattle) were 1) to estimate genetic parameters for direct and maternal genetic effects for birth weight, weaning weight, and six months weight which can be used for genetic evaluations and 2) to compare models with and without grandmaternal effects. Data were obtained from the National Livestock Research Institute in Rural Development Administration (RDA) of Korea and were used to estimate genetic parameters for birth weight (BW, n= 10,889), weaning weight at 120-d (WW, n=8,637), and six month weight (W6, n=8,478) in Hanwoo. Total number of animals in pedigrees was 14,949. A single-trait animal model was initially used to obtain starting values for multiple-trait animal models. Estimates of genetic parameters were obtained with MTDFREML using animal models and derivative-free REML (Boldman et al., 1995). Estimates of direct heritability for BW, WW, and W6 analyzed as single-traits were 0.09, 0.03, and 0.02 from Model 3 which included direct and maternal genetic, maternal pemianental environmental effects, and effects due to sire X region x year-season interaction, respectively. Ignoring sire x region x year-season interaction effect in the model (Model 2) resulted in larger estimates for direct heritability than for Model 3.
Estimates of maternal heritability for BW, WW and W6 were 0.04, 0.05, and 0.07 from Model 3, respectively. The estimates of direct-maternal genetic correlation were positive for BW, WW, and W6 with Model 3 but were negative with Model 2 for WW and W6. Estimates of direct genetic correlations between BW and WW, BW and W6, and WW and W6 were large: 0.52, 0.45, and 0.90, respectively. Genetic correlations were also large and positive for maternal effects for BW with maternal effects for WW and W6 (0.69 and 0.74), and even larger for WW with W6 (0.97). The log likelihood values were the same for models including grandmaternal effects as for models including maternal effects for all traits. These results indicate that grandmaternal effects are not important for these traits for Hanwoo or that the data structure was not adequate for estimating parameters for a grandmaternal model. (Asian-Aus, J. Anim. Set 2000. Vol. 13, No. 2 : 149-154) Key Words : Hanwoo, Genetic Parameters, Maternal, Model
INTRODUCTION
Genetic parameters for weight traits at birth and weaning for Korean Cattle which is called Hanwoo using models including maternal effects have not been previously reported extensively. Maternal effects may be important to offspring for some traits. Improvement in weaning weights of beef calves has been found to depend on increased preweaning growth potential of calves and maternal ability of cows (e.g., Mangus and Brinks, 1971). Birth weight and growth rate have been shown to be influenced by an animaFs genetic potential and by the maternal environment (Meyer, 1992). Numerous studies of preweaning growth of beef cattle have reported a genetic correlation between direct and maternal genetic effects (Bertrand and Benyshek, 1987; Brown and Galvez, 1969; Burfening
et 이., 1981; Garrick et al., 1989; Koch, 1972; Kriese et al., 1991; Nelsen et al., 1984; Yokoi et al., 1997).
Willham (1963) suggested that the maternal effect of the granddam may influence the maternal effect of the dam. He (1972) developed principles for biometrical analyses of direct and maternal effects and again proposed the possible influence of a grand- maternal effect. Dodenhoff et al. (1998) reported that grandmaternal effects were not important for birth weight but seemed to be important for weaning weight in Herefords. They suggested that when the model did not include grandmaternal effects, maternal heritability may be underestimated.
Models which include the effects of sire Xherd X year interaction (Dodenhoff et al., 1999) and of sireX year interaction (Lee and Pollak, 1997; Robinson,
1994) in beef cattle have been investigated. Lee and Pollak (1997) and Robinson (1994) reported that the direct-maternal genetic correlation increased when the effects of sire X year interaction were included in the model.
The objectives of this study of Hanwoo were 1) to estimate genetic parameters for direct and maternal genetic effects with and without effects in the model for effects of sire Xregion xyear-season interaction, 2) to compare models with and without grandmaternal
150 CHOI ET AL.
effects, and 3) to estimate genetic correlations among the traits.
MATERIALS AND METHODS
The materials and methods section is divided into two parts. The first part is a description of direct and maternal genetic effects included in the model. The second part is a description of grandmatemal genetic effects included in the model. Only records with maternal granddam known were used in the second part.
Section I. Model with direct and maternal genetic effects
Data were collected from 1974 through 1995 by the National Livestock Research Institute in Rural Development Administration (RDA) of Korea and were used to estimate genetic parameters for birth weight (BW, n=10,889), 120-d weaning weight (WW, n=8,637), and 180-d six month weight (W6, n=8,478).
Total number of animals in pedigrees was 14,949.
Numbers of records, overall means, and standard deviations by trait and by sex are in table 1. Adjusted weaning weight at 120-d was calculated as:
actual weaning
Adjusted weight-birth weight birth 120-day = ---X120 + . K
weight age in days weight
at weaning
Three single-trait animal models were initially used for this analysis. Then estimates from the most
appropriate sin이e trait model were used as starting values for two-trait analyses. RegionX year-season and age of dam X sex of calf were fitted as fixed effects for all models. The spring season is centered on April and the fall season is centered on October. Two recording regions were involved. Herd identification was unavailable. Animals were intact males and females. Thus, steers were not included in these data.
Table 2 presents genetic parameters to be estimated with the three models.
Table 2. Parameters estimated with the three models
Model Parameters3
房
am, 0"；爲
1 2b
3b
、 /
a direct genetic variance, 房”느 maternal genetic variance, oam= covariance between direct and maternal genetic effects, o%= variance of maternal permanent environmental effects, 兄드 variance of sire X region X year-season interaction effects, and
夜
= variance of temporary environmental effects.b If sire identification was missing then the record was deleted.
Model 1 included direct and maternal genetic, maternal peimanental environmental and residual environmental effects,
y=XB +Z]d+Z 기n+W 也+。, where
y = the NX1 vector of observations,
'Records unadjusted for the model effects.
b BW = birth weight (kg), WW = weaning weight (kg), and W6 = six month weight (kg).
Table 1. Number of observations, overall means (kg)a including progeny with missing sire identification
and standard deviations (SD) by trait including and not
Trait0 No. of records No. of sires No. of dams No. of SRYSC mean ± SD
BW 10,889 1,386 3,908 - 23.02± 2.76
WW 8,637 1,300 3,423 - 95.39 ±20.55
W6 8,478 1,243 3,419 - 131.76± 19.36
l^iUL UIVlUUlllg UJUQSLL 丄g 3U3
BW Overall 10,747 1,346 3,826 1,506 23.01 ± 2.76
Female 5,207 946 2,619 1,086 22.20 ± 2.52
Male 5,540 885 2,607 1,012 23.77 士 2.76
WW Overall 8,528 1,263 3,355 1,396 95.41 ±20.49
Female 3,929 852 2,244 965 91.03± 18.88
Male 4,599 858 2,300 969 99.15±21.07
W6 Overall 8,377 1,208 3,359 1,347 131.78± 19.35
Female 3,696 765 2,115 881 126.04±16.14
Male 4,681 875 2,396 990 136.30±20.45
SRYS = sire X region X year-season interaction.
B = vector of fixed effects (regionxyear-season and age of dam xsex combinations),
a = vector of breeding values for direct genetic effects, m = vector of breeding values for maternal genetic
effects,
p = vector of maternal permanent environmental effects, e = vector of random residual effects, and X, Zi, Z爲 and Wi are known matrices relating observations in y to fixed and random effects.
Model 2 was based on Model 1 but if identification of a sire was missing then the record was deleted. Model 3 was based on Model 2 but was extended to include effects due to sire xregion x year-season (SRYS) interaction,
y=X B +Zia+Z2m+Wijp+W2^+e>
where for the added part of the model q = vector of sirex region X year-season interaction effects, and Wi is a known matrix relating observations in y to random effects in q.
For the model:
E [y]=",
and the (co)variance structure of the random effects for Model 3 is;
a 「A，土 ” 2
u a A（丁 am 0 0 0
m •A•(丁 am A.（丁 m 0 0 0
P = 0 0 -*Nd 6 p2 0 0
q 0 0 0 JNs 0 q2 0
e 0. 0 0 0 /N (7 e _2
where
Nd = number of dams,
Ns = number of sire x region x year-season combinations, N = number of records,
A = numerator relationship matrix among animals in the pedigree file, and
I = identity matrix of appropriate order.
Estimates of genetic parameters were obtained with MTDFREML wliich is a set of programs for estimating (co)variance components using animal models and derivative-free REML (Boldman et 시., 1995).
Section II. Model with direct, maternal and grand- maternal genetic effects
Data were birth (BW, n=5,023), 120-d weaning (WW, n드4,218), and 180-d six month weights (W6, n=4,040). All animals with records were required to have a known granddam. Total nxrnber of animals in
pedigrees was 7,785. Three single-trait animal models were compared. Region xyear-season and age of dam xsex combinations were fitted as fixed effects for all models. However, effects of sirex regionx year-season (SRYS) interaction were not included in the model in this section.
Effects in Model 1 included direct and maternal genetic, maternal permanent environmental, and residual environmental effects.
Model 2 was based on Model 1 but was extended to include a vector of grandmatemal effects assumed to be uncorrelated.
Model 3 was based on Model 2 but was extended to include grandmatemal genetic effects.
Estimates of genetic parameters were obtained with REML using the average information method (Al- REML) (Johnson et 시., 1995) with a program developed by Dodenhoff et al. (1998).
The results were inconclusive as estimates (and likelihoods) were the same as for the maternal effects model. The most likely reason is that the data structure was inadequate for estimating the six genetic (co)variance components and the three environmental (co)variance components. A less likely reason is that grandmatemal effects are nil for Hanwoo cattle.
RESULTS AND DISCUSSION
For all single trait analyses, estimates with Models 1 and 2 were essentially the same as only a few animals did not have an identified sire, Therefore this section will compare results with Models 2 and 3.
Birth weight
Parameter estimates for birth weight using the three models are shown in table 3 for single trait analyses.
Estimates of direct and maternal heritability with Model 2, which ignored effects due to sirex region x year-season (SRYS) interaction were 0.14 and 0.04, respectively. Estimates of direct and maternal heritability with Model 3, which included effects due to sireX regionx year-season (SRYS) interaction were 0.09 and 0.04, respectively. Failure to include sire x regionx year-season interaction effects in the model (Model 2) resulted in a higher estimate for direct heritability than for Model 3. The direct-maternal genetic correlation with Model 3 was large and positive (0.61) but was only 0.23 with Model 2. This result agrees with Meyer (1992) who reported a positive direct-maternal genetic correlation of 0.29 in Australian beef cattle.
Including effects due to sire x regionx year-season interaction (Model 3) resulted in an improvement in the log likelihood compared to Model 2 and estimates of variance components with Model 3 were significantly different from those for Model 2. The estimate of direct heritability was less and direct-
152 CHOI ET AL.
Table 3. Estimates of variance components and genetic parameters (asymptotic standard errors) for birth, weaning and six month weights
a Birth weight Weaning weight Six month weight
Parameters ___________________________ _________________________________________________________
Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 -21og L 29336.74 28925.48 28920.37 54935.06 54198.86 54187.91 55632.92 54971.96 54956.57
死
0.82 0.77 0.48 31.42 30.29 7.46 39.42 38.81 6.21° m2 0.25 0.25 0.23 17.04 16.95 11.88 28.30 29.09 20.99
am 0.09 0.10 0.20 -10.31 -9.87 1.08 -11.60 -11.76 4.53
0.16 0.14 0.13 25.72 24.85 25.02 29.54 28.68 29.04
。匕
0.09 7.24 9.74196.65 197.65 211.46 163.91 164.04 173.67
o.
(0.
0.
(0.
0.
(0.
0.
(0.
0.
3 3)o 4)58)o 3) X1 1tu X nu
cli
X1 1X 4
3)7 4)5
"o 2) 1 0 0 0 3 1 0
■
•
>
•
•
♦
■ o (00 (0 -0 (0.0 (o
OJ O.
z(x 70
°- 70
°- 2) (o.
o 3
4 - 0 3 1 2 1 2 3 1 7 2 0 0 0 0 1 7 1 0 0 0 7 0 o.
(0.
0.
(0.
0.
(0.
0.
(0.
0.
(0.
0.
(o.
3 3 7 3 4 7 1 2 1^ 00 0^ 4^ 11 0^ , Q Q z(x z(x - z(
x /—、
°-
3 2) 7 o XI/
XI/
XI/
XI/
4 3 7 3 5 4 1 2 1^ 00
^0 4^ 31
^0
。
z(x z(x _ z(
x /—、
2 2) 7 o 9
3)4 2)18)2 2)2 Do 2) o o o o 6 4 o o o o 8 o o o o o o o o o o z(x z(x z(x z(x
-( -J o o o xz(
/(
XI/
XI/
XI/
XI/
2 4 2 4 2 3 6 3 2 31 -0 0^ 07
^ 20 0^
。
z(x z(x z(x z(x 9 2)45 2)o9)3 2) 2 2 0 0 0 2 1 0 0 z— x z(x z(x X/I
76 02
°- o幻
4 4
7 2 7 o Q y am
护
o\= direct genetic variance, <?《= maternal genetic variance,
(
y(inj= covariance between direct and maternal effects, variance of maternal permanent environmental effects, 兄=variance of sire X region X year-season interaction effects, o2=variance of temporary environmental effects, h\= direct heritability, 勿%= maternal heritability, 尸血드 genetic correlation between direct and maternal effects, p2= fraction of variance due to maternal permanent environmental effects, q2=
fraction of variance due to sire XregionXyear-season interaction effects, and e2= fraction of variance due to temporary environmental effects.
maternal genetic correlation were greater with Model 3 than with Model 2. Earlier studies showed mostly negative correlations between direct and maternal effects for birth weight as reviewed by Mohiuddin (1993). This analysis indicates that inclusion of sireX regionXyear-season interaction effect in the model seems to be important for birth weight for Korean Cattle.
Weaning weight
Parameter estimates for weaning weight with the three models also are shown in table 3. The pattern for estimates for weaning weight is similar to the pattern for birth weight. Estimates of direct and matemal heritability with Model 3 were 0.03, and 0.05, respectively compared to 0.13 and 0.07 with Model 2. The direct-maternal genetic correlation from Model 3 was positive (0.11) but was negative (-0.44) with Model 2. The estimate of 0.11 agrees with Wright et al. (1987) who reported a positive direct- maternal genetic correlation of 0.16 in American Simmental cattle. Inclusion of effects due to sire X regionx year-season interaction (Model 3) resulted in an improvement in the log likelihood compared to Model 2. Estimates of variance components for Model 3 were significantly different from those from Models
2 and 1 because including effects of sire x region x year-season interaction increased the direct-maternal genetic correlation from large negative (-0.44) to slightly positive (0.11) and also reduced the estimate of direct heritability. This pattern has been reported by Robinson (1994) and by Lee and Pollak (1997).
Numerous studies of beef cattle have found mostly negative genetic correlation correlations between direct and maternal effects for weaning weight as reviewed by Mohiuddin (1993). Model 2 that ignored sireX region Xyear-season interaction for weaning weight resulted in a significantly poorer log likelihood. The results indicate that inclusion of sire X regionX year- season interaction effects in the model for weaning weight seems to be important for Korean Cattle.
Six month weight
Parameter estimates for six month weight with the three models are presented in table 3. The pattern for estimates for six month weight is similar to the pattern for birth and weaning weights. Estimates of direct and maternal heritability with Model 3 for six month weight analyzed as single-traits were 0.02, and 0.07, respectively but were 0.13 and 0.10 with Model 2. Including effects of sire X region X year-season inter
action also increased the direct-maternal genetic
con-elation from large negative (-0.35) to positive (0.40) and also reduced the estimate of direct lieritability compared to Model 2. The difference in log likelihoods between Models 2 and 3 was significant indicating including effects of sire X region Xyear-season interaction in the model is important.
The differences in estimates between Models 2 and 3 follow the same pattern for W6 as for WW as would be expected. The results indicate that inclusion of effects of sire X region Xyear-season interaction for six month weight in the model also is important for analyses of six month weight in Hanwoo.
With these early-in-life traits, positive estimates of genetic coiTelation between direct and maternal genetic effects were found when sire X region Xyear-season interaction effects, were included in the model (Model 3). The model ignoring effects of sire Xregion X year-season interaction resulted in negative genetic correlations between direct and maternal genetic effects except for birth weight. For all traits, including the interaction effects resulted in an increase in the direct-maternal genetic correlation and a decrease in direct heritability. These results agree with those of Robinson (1994) and Lee and Pollak (1997) who reported that estimates of direct-maternal correlations increased when sire X year effects were included in the model.
Multiple trait analyses
Parameter estimates for two-trait analyses for birth, weaning, and six month weights are shown in table 4.
Model 3, which included effects of sire X region X year-season (SRYS) interaction, was the best fitting model for sin이e trait analyses of birth, weaning, and six month weights. Therefore, Model 3 was used for the two-trait analyses.
Estimates of direct genetic correlations between BW and WW, and between BW and W6 were 0.52 and 0.45, respectively which are moderate and positive, and between WW and W6 was large (0.90).
Estimates of direct with maternal genetic correlations were similar to those from the single trait analyses with Model 3. Estimate of genetic correlation between direct genetic for BW and maternal effects for WW was near zero, -0.04, The estimate of the maternal genetic correlation between BW and WW was 0.66.
Estimate of maternal permanent environmental correlation between BW and WW was 0.20 and between WW and W6 was large (0.99). Estimates of environmental con'elations between BW and WW and between BW and W6 were 0.17 and 0.17, respectively, the estimate between WW and W6 was large (0.73). Nelsen and Kress (1979) reported an estimate of direct genetic correlation between BW and WW to be 0.53 for Angus in USA. Koots et al. (1994) in a review reported the average of estimates of direct genetic
Table 4. Parameter estimates from two-trait analyses for birth and weaning weight, birth and six month weights, and weaning and six month weights
1 BW BW WW
2 WW W6 W6
5.59 5.59 234.39
成 227.81 283.73 312.50
片 0.08 0.08 0.06
hl 0.04 0.03 0.05
hl, 0.05 0.05 0.05
後 0.06 0.09 0.07
r f>\ 0.52 0.45 0.90
0.59 0.54 0.11
戶 <7( "E -0.04 0.21 0.19
y m. 0.66 0.83 0.17
r<7; JWj 0.03 0.26 0.32
0.69 0.74 0.97
a2
p\ 0.02 0.01 0.11
p
!
0.11 0.10 0.110.02 0.02 0.03
展 0.03 0.04 0.03
r Plh 0.20 0.00 0.99
r <7l S 0.59 0.73 0.87
屛 0.79 0.79 0.75
0.76 0.74 0.72
0.17 0.69 0.73
BW= birth weight, WW= weaning weight, W6= six month weight, g
}=
phenotypic variance for trait i, h%=direct heritability for trait i, 肉二=maternal heritability for trait , r 그 genetic correlation between direct effects for trait i and j, r genetic correlation between direct and maternal effects for trait i, r (f.m= genetic correlation between direct effects for trait i and maternal effects for trait j, r a.m= genetic correlation between direct and maternal effects for trait j, and 2,力>=fraction of variance due to maternal permanent environmental effects for trait i, 시 = fraction of variance due to sire X region X year-season interaction effects for trait i, r correlation between maternal permanent environmental effects for trait i and j, correlation between sire X region X year-season interaction effects for trait i and j, and
以
=fraction of variance due to temporary environmental effects for trait i, and r e,e= correlation between temporary environmental effects for traits i and j.
genetic correlation between BW and WW to be 0.50, the average of estimates of correlation between direct genetic for BW and maternal effects for WW to be -0.14, the average of estimates of correlations between direct genetic for WW and maternal effects for BW to be -0.05, and the average of estimates of maternal genetic correlations between BW and WW to be 0.39.
154 CHOI ET AL.
CONCLUSIONS
Results of this study suggest that maternal effects are important for Hanwoo (Korean Native Cattle) for birth, weaning, and six month weights. Including sire
XregionXyear-season interaction effects in the model for genetic analysis of birth, weaning, and six month weight may be important because interaction effects seem to influence variance and covariance components associated with direct and maternal genetic effects. The data sti'ucture was not sufficient to determine whether grandinatemal effects are important for birth, weaning, and six month weight for Hanwoo. The large genetic coiTelations between weights at weaning (about four months) and at six months and similar estimates of heritabililty indicate little reason to maintain or analyse both weaning and six-month weights in Hanwoo.
REFERENCES
Bertrand, J. K. and L. L. Benyshek. 1987. Variance and covariance estimates for maternally influenced beef growth traits. J. Anim. Sci. 64:728-734.
Boldman, K. G., L. A. Kriese, L. D. Van Vleck, C. P. Van Tassell and S. D. Kachman. 1995. A manual for use of MTDFREML. A set of programs to obtain estimates of
variances and covariances. USDA, ARS.
Brown, C. J. and V. Galvez. 1969. Maternal and other effects on biilh weight of beef calves. J. Anim. Sci. 28:162-167.
Burfening, P. J., D. D. Kress and R. L. Friedrich. 1981.
Calving ease and growth rate of Simmental sired calves.
III. Direct and maternal effects. J. Anim. Sci. 53:1210- 1216.
Dodenhoff, J., L. D. Van Vleck and D. E. Wilson. 1999.
Comparison of models to estimate genetic effects for weaning weight of Angus cattle. J. Anim. Sci. (in press).
Dodenhoff, J., L. D. Van Vleck, S. D. Kachman and R. M.
Koch. 1998. Parameter estimates for direct, maternal, and grandmatemal genetic effects for birth weight and weaning weight in Hereford cattle. J. Anim. Sci. 76:
2521-2527.
Ganick, D. J., E. J. Pollak, R. L. Quaas and L. D. Van Vleck. 1989. Variance heterogeneity in direct and matemal weight traits by sex and percent purebred for Simmentaksired calves. J. Anim. Sci. 67:2515-2528.
Johnson, D. L. and R. Thompson. 1995. Restricted maximum likelihood estimation of variance components for univariate animal models using sparse matrix techniques and average
information. J. Dairy Sci. 78:449-456.
Koch, R. M. 1972. The role of matemal effects in animal breeding. VI. Matemal effects in beef cattle. J. Anim.
Sci. 35:1316-1323.
Koots, K. R., J. P. Gibson and J. W. Wilton. 1994. Analyses of published genetic parameter estimates for beef production traits. 2. Phenotypic and genetic correlations.
Anim. Breed. Abst. 62:825-853.
Kriese, L. A., J. K. Bertrand and L. L. Beny아］iek. 1991.
Genetic and environmental growth trait parameter estimates for Brahman and Brahman-derivative cattle. J.
Anim. Sci. 69:2362-2370.
Lee, C. and E. J. Pollak. 1997. Relationship between sire X year interactions and direct-maternal genetic correlation for weaning weight of Simmental cattle. J. Anim. Sci.
75:68-75.
Mangus, W. L. and J. S. Brink. 1971. Relationships between direct and maternal effects on growth in Herefords. I.
Environmental factors during preweaning growth. J.
Anim. Sci. 32:17-25.
Nelsen, T. C. and D. D. Kress. 1979. Estimates of heritabilities and correlations for production cahracters of Angus and Hereford calves. J. Anim. Sci. 48:286-292.
Nelsen, T. C., R. E. Short, J. J. Urick and W. L. Reynolds.
1984. Genetic variance components of birth weight in a herd of unselected cattle. J. Anim. Sci. 59:1459-1466.
Meyer, K. 1992. Variance components due to direct and matemal effects for growth traits of Australian beef cattle. Livest. Prod. Sci. 31:179-204.
Mohiuddin, G. 1993. Estimates of genetic and phenotypic parameters of some performance traits in beef cattle.
Anim. Breed. Abst. 61:495-522.
Robinson, D. L. 1994. Models which might explain negative correlations between direct and maternal genetic effects.
5th World Congr. Genet. Appl. Livest. Prod. 18:378-380.
Willham, R. L. 1963. The covariance between relatives for characters composed of components contributed by related individuals. Biometrics. 19:18-27.
Willham, R. L. 1972. The role of maternal effects in animal breeding: III. Biometrical aspects of maternal effects in animals. J. Anim. Sci. 35:1288-1293.
Wright, H. B., E. J. Pollak and R. L. Quaas. 1987. Estimation of variance and covariance components to determine heritabilities and repeatability of weaning weight in American Simmental cattle. J. Anim. Sci. 65:975-981.
Yokoi, N., K. Moriya and Y. Sasaki. 1997. A measure for predicting genetic merit for milking and nursing ability in beef cattle. Anim. Sci. 65:39-43.
